Puzzle Solving: Wrong answer
Puzzle Solving
Wrong answer
See which AI models are most likely to hit Wrong answer on Puzzle Solving, so you can spot weak points faster. Sort by: Tests Correct ↑.
Failure Reasons
246/246
Filter models
No models match the current search and filters.
| Rank | Model | Company | Wrong answer Count | Category Score | Total Cost | Tests Correct | Response Time (avg) |
|---|---|---|---|---|---|---|---|
| #92 | Claude Haiku 5.5 high | Anthropic | 1 | 0.6 | $0.088 | 1/3 | 3.28s |
| #100 | DeepSeek V4.1 Flash 1017 max | DeepSeek | 1 | 0.6 | $0.757 | 1/3 | 12.3s |
| #106 | DeepSeek V4 Pro 0423 high | DeepSeek | 1 | 0.7 | $0.149 | 1/3 | 56.8s |
| #111 | Nemotron 3 Ultra medium | NVIDIA | 2 | 0.5 | $0.674 | 1/3 | 3.54s |
| #126 | DeepSeek V4.1 Flash 1017 high | DeepSeek | 1 | 0.5 | $0.589 | 1/3 | 5.70s |
| #129 | Step 3.7 Flash medium | Stepfun | 2 | 0.6 | $0.571 | 1/3 | 6.19s |
| #136 | Muse Glimmer 30B high | Meta | 1 | 0.6 | $0.506 | 1/3 | 64.8s |
| #140 | Qwen3.7 Flash low | Qwen | 1 | 0.7 | $0.071 | 1/3 | 9.79s |
| #141 | Muse Glimmer 30B medium | Meta | 2 | 0.6 | $0.288 | 1/3 | 64.7s |
| #149 | Mistral Large 4 high | Mistral | 2 | 0.5 | $1.391 | 1/3 | 55.2s |
| #168 | Qwen3.5-27B none | Qwen | 1 | 0.7 | $0.116 | 1/3 | 1.38s |
| #173 | Qwen3.7 Flash high | Qwen | 1 | 0.7 | $0.063 | 1/3 | 10.8s |
| #174 | LongCat 2.0 medium | Meituan | 2 | 0.5 | $0.767 | 1/3 | 8.84s |
| #176 | Step 3.7 Flash low | Stepfun | 2 | 0.6 | $0.449 | 1/3 | 1.84s |
| #187 | Kimi K2.7 Code medium | Moonshot AI | 1 | 0.6 | $1.136 | 1/3 | 41.0s |